COLOR CONFINEMENT, ABELIAN GAUGE AND RENORMALIZATION-GROUP FLOW

Under the assumption that the color charge can be written in a BRST exact form, the color confinement mechanism proposed by Kugo and Ojima (KO) explains the confinement of any colored particles including dynamical quarks and gluons. This mechanism, however, is known to break down in the abelian gauge which treats the maximal abelian subgroup of the gauge group in a special manner. In order to study whether the failure of the KO mechanism is particular only to the abelian gauge or whether this failure occurs in a wide class of gauges including the ordinary Lorentz type gauge, we carry out a renormalization group study of the SU(2) gauge theory in the gauge fixing space. Our gauge fixing space consists of four distinct regions that are not connected with each other by renormalization group flows, and we find that the abelian gauge is infrared unstable in three regions which include the Lorentz type gauge. This suggests that the failure of the KO mechanism is a phenomenon which occurs only in the abelian gauge. We also find that the Lorentz gauge is infrared stable.